Measuring Cell Type Similarity with Gene Ontology in Single-Cell RNA-Seq

Traditional methods for analyzing single cell RNA-seq datasets focus solely on gene expression, but this package introduces a novel approach that goes beyond this limitation. Using Gene Ontology terms as features, the package allows for the functional profile of cell populations, and comparison within and between datasets from the same or different species. Our approach enables the discovery of previously unrecognized functional similarities and differences between cell types and has demonstrated success in identifying cell types' functional correspondence even between evolutionarily distant species.

scGOclust

Leveraging Gene Ontology to Measure Cell Type Similarity Between Single Cell RNA-Seq Datasets

Author & Maintainer: Yuyao Song [email protected]

Note main branch version is compatible with Seurat and SeuratObject >= V5.0. If you are still using V4.0, please go to release scGOclust_V0.1.3.

Installation

1. First create the conda environment. Mamba is recommended as a faster alternative for conda.

   conda env create -f scGOclust_conda_7Dec2022.yml

2. Then, open R under this environment, and install several packages not in conda:

   remotes::install_github('satijalab/seurat-wrappers'), install.packages("pheatmap", "slanter")

3. Finally, install scGOclust from GitHub:

   devtools::install_github("Papatheodorou-Group/scGOclust", ref = "main")

Usage

This package operates on pairs of Seurat objects

Refer to vignettes for usage examples

License: GPLv3.0